摘要
椭球单元通过高斯分布逼近形成各模式类的决策区域，是一种非常适合于模式识别任务的前馈型人工神经网络模型。提出改进椭球单元神经网络的训练权重组，即采用多权重组，增强了椭球单元网络的抗干扰能力，提高了网络的故障诊断能力，并给出了权重选择方法。仿真和试验验证了该方法的正确性。
Neural Networks with Ellipsoidal Activation Functions closes in upon a decision making region by Gauss distribution for various patterns and is adapted to fault diagnosis well. It has the advantages, such as high precision in classifying, fast training rate and good rejective for unknown pattern. For improving the anti jamming performance, a multi weight is adopted between input neuron and ellipsoidal neuron in this paper. The selecting method of weights is also derived for network training. The simulation and experiments sustain the multi-weight method.
出处
《中国机械工程》
EI
CAS
CSCD
北大核心
1999年第8期890-893,共4页
China Mechanical Engineering
基金
国家自然科学基金
关键词
椭球单元
神经网络
故障诊断
噪声
ellipsoidal activation functions neural network fault diagnosis noise
